Chris takes up the question that several commenters raised on dzone, i.e. how do we define ‘lightweight’ when considering whether Spring still is. He digs back to Rod Johnson’s original mission statement for Spring, and concludes that Spring still does meet those criteria.
I think he’s right. What I was whining about in my post was really not that Spring, as a framework, is no longer a lightweight platform. It is, and is still my starting point when building an app of more than trivial size.
My moaning was really about the ecosystem of enterprise solutions that Spring Source has built up around the framework. Trying to figure out what these various things actually are is a pain. There is a lot of useful stuff there, but it requires a pretty heavy investment, in both time and cash, to work out what it all is, which parts may be useful to my organization, and how to fit them into our architecture.
On the plus side, it’s all still open source at it’s core, so once we do work out what’s what, we have plenty of choice about how to use it. If our needs are more complex, and our budget larger, it’s worth looking at splashing out for the enterprisey, supported versions that Spring Source offers. Having played with Roo, and investigated Spring Integration and other web services pieces, it’s clear they’re still a pragmatic, dare I say it, lightweight alternative to the stuff put out by the likes of IBM, HP, and Oracle, and one that’s still acceptable to pointy-haired bosses.
What’s truly key is that if our technical strategy favours the use of lighter, open source components, the Spring folks are still putting out stuff that meets our needs. Hopefully they won’t lose sight of that.